Replying to Avatar HoloKat

A few reason not to have algorithms in nostr:

1. Higher signal from people you follow.

Without algos, you have to be selective about who you follow, and you’re more likely to curate that group of people often. Since content is not surfaced via some relational query, you just see what your circle posts.

Nostr without algorithms also resembles life more closely than other social networks. You have to literally be online to see what someone said. Sleeping? You may miss their note unless it is not buried enough that you can go back to it, or if they mention you directly. I think anything that more closely resembles real life interactions is ultimately healthier for us.

Of course that doesn’t mean you’ll miss everything. If the note is great, enough people will like or zap it that sorting can still surface it for you when you are back.

2. Better relationships from smaller number of direct connections. If the Dunbar’s number is to be trusted, we are limited at 150 people with whom we can maintain social connections. Without algos and curation of people you follow, you are more likely to form stronger connections. This in turn makes the network feel more intimate and friendly. Even as we speak I follow 178 people and it feels just about right.

3. Ability to step away without being punished. At least on Twitter, anecdotal evidence suggests that if you step away, your reach diminishes and you have to “earn” it back. Without algos on Nostr you don’t need to worry about this. Just step away as long as you need to - better overall for mental health. Come back when you like and you’ll still reach everyone who follows you. This is exactly how real life works - you enter a room of people and they will all hear you instead of a tiny fraction.

While other social networks optimize for engagement and time spent on website, I think nostr should optimize for maximum mental health and least stress.

Curious to hear ya’ll thoughts…

The main reason I will consider algos is to discover content. I see Nostr as far more than just a decentralized twitter, possibilities are endless.

So, I agree with you, in my Twitter like client I am fine without algos. I just want my lil tribe to face the future with. But, if I am browsing for content (blogs, books, music, videos, research papers…) I foresee that an algo I can control could help me finding more stuff that is of interest to me.

Reply to this note

Please Login to reply.

Discussion

Better than standard sorting? (Most liked, most zapped; liked by friends, etc…)?

Possibly, it’s all possibilities for now. I am curious to see what algos could be capable of doing if they are coded by people with good intentions.

Decentralizing algos development would surely provide some interesting outcomes I believe.

I think it would only centralize over time and go back to exactly what some seek to escape

Do you have the same fears for #nostr as a whole?

Not as much, no, because we’ll have so many clients to choose from.

So why wouldn’t that be the case with algos? We could have many algos to pick from, or even build our own combining predefined blocks from algos service providers.

Well, there’s no way to stop anyone from adding algos to clients. I’m just making a case for why you should probably think about it first…

Fair point. Let’s hope to have clients and relays who will leave that choice to the user. If that is possible.

I am not a coder, so when I browse the NIPs I might get confused. I thought algos would be something a user could control.

Sort by zaps , sort by reactions within, sort by comments and new. And also set time period (default 24hours)

I think these four filters/sort should suffice discovery

I feel that is a waste of all the tags and metadata that will be available. But, obviously if I cannot chose/control the algo, I do not want it.

Even in its simplest form, I could design an algo that gives parameters for all the things you have listed and combine zaps, retweet, likes, comments and more in one feed you configure to your linking.

Similar to Reddit *Hot* section